Method and engine system for reducing cylinder air-fuel ratio imbalance

Abstract

Procedure comprising: Sampling an engine speed signal concurrent with engine firing; Filtering the sampled engine speed signal via a notch filter; and Indicating a cylinder air / fuel imbalance based on an output of the notch filter with respect to a threshold value, further comprising adjusting one or more engine operating parameters in response to the indication of the cylinder air / fuel imbalance.

Description

Area

[0001] The present description generally relates to methods and systems for detecting and treating air-fuel imbalances in a cylinder based on crankshaft acceleration signals. State of the art / brief description

[0002] Cylinder-to-cylinder combustion variations associated with air-fuel ratio imbalances can occur in engines for a variety of reasons. For example, cylinder-to-cylinder air-fuel ratio imbalances can occur due to intake valve deposits, clogged exhaust gas recirculation (EGR) bores, electrical problems, air leaks, and / or misaligned fuel injectors. When an air-fuel ratio imbalance occurs in one or more cylinders, engine performance degrades. Additionally, an engine may be unable to meet emissions compliance, and fuel economy may be reduced.

[0003] From DE 42 27 104 A1 a method is known whereby a torque irregularity or an air / fuel imbalance in an engine cylinder can be detected based on a sampled and filtered engine speed signal.

[0004] From DE 10 2013 200 323 A1 it is known to identify an air / fuel imbalance in a cylinder of an engine based on the evaluation of crankshaft acceleration signals.

[0005] An exemplary approach for detecting air-fuel ratio imbalance is shown by Javaherian in US Pat. No. 6,668,812 B2. Therein, a time-sequential series of signals from an exhaust oxygen sensor is collected over at least one engine cycle at a current engine speed and current load conditions, and the series of signals is converted by a discrete Fourier transform into a vector of air-fuel ratio imbalances at a specific frequency. The vector is then projected onto two fuel imbalance reference vectors of known magnitude and phase, corresponding to the discrete Fourier transform of two nominal fuel imbalance patterns for the current engine speed and load. The reference vectors are previously calibrated and stored in the memory of an engine controller.An air-fuel imbalance in a cylinder is detected based on a deviation of the sampled vector from the reference vector.

[0006] However, the present inventors have recognized a potential problem with such systems. Detecting air-fuel imbalances using the method of US Pat. No. 6,668,812 B2 can be time-consuming, costly, and computationally intensive due to the high sampling rates involved and the complexity of the required vector transformation. In particular, relying on the fast Fourier transform (FFT) can lead to delays in detecting and handling air-fuel ratio imbalances.

[0007] The present inventors have recognized the above-mentioned problems and identified an approach to at least partially address the above-mentioned problems. In one example, the above-described problems may be addressed by a method for normalizing engine speed content at a selected frequency with respect to changes in crankshaft angle by sampling engine speed concurrently with engine firing events, processing the sampled engine speeds using a notch filter at the selected frequency, and detecting cylinder imbalance based on the normalized sampled signal. This allows analysis across a frequency range of sampled engine speed values ​​using a notch filter, simplifying cylinder imbalance detection without compromising the reliability of the results.

[0008] As an example, a consecutive series of sampled engine speed values ​​may be collected within one engine cycle of an engine where all cylinders are to be controlled with stoichiometry. The sampled values ​​may then be filtered using a discrete notch filter set to a frequency of once per engine cycle. Additionally, before processing the sampled values ​​with a notch filter, the sampled values ​​may be normalized for changes in crankshaft angle and stored in the memory of an engine controller. Therein, engine speed signals are sampled during the power stroke from each firing cylinder, and the estimated acceleration based on the torque from each cylinder is normalized by scaling based on the deviation of the ignition timing from the MBT spark.To detect the presence of an air / fuel imbalance in an engine cylinder, the engine speed content may be sampled at multiple locations within an engine firing event. The sampling frequency may be set based on the engine configuration and the cylinder firing frequency, and in one example, is an integer multiple of an engine firing frequency. The sampled data is then processed using a discrete notch filter set to the sampling frequency and with values ​​to cancel a frequency of once per engine cycle to obtain a frequency-domain characterization of the engine speed. The notch filter output is subtracted from the original signal. The processed output level is then compared to a threshold level. Based on the level relative to the threshold level, the presence of a cylinder imbalance may be determined.Additionally, the degree of imbalance and the alignment of the imbalance (i.e., whether the imbalance is richer or leaner than stoichiometry) can also be determined based on the level and phase of the processed output. Engine parameters can then be adjusted to reduce the imbalance. For example, the fuel supply to the unbalanced cylinder can be adjusted to correct the imbalance.

[0009] This allows for inter-cylinder changes in the air / fuel ratio to be monitored. The technical effect of employing a discrete notch filter for frequency-domain characterization of engine speed content is that air / fuel ratio imbalances can be detected using faster and simpler processing techniques. In particular, the need for complex, time-consuming, and computationally intensive processing techniques, such as the Fourier transform, is reduced without reducing the accuracy of air / fuel ratio imbalance detection. Overall, by detecting a cylinder's air / fuel imbalance with greater reliability, emissions can be reduced and engine performance can be improved.

[0010] It should be understood that the above summary is provided to introduce, in simplified form, a selection of concepts further described in the detailed description. It is not intended to identify key or essential features of the claimed subject matter, the scope of which is defined solely by the claims following the detailed description. Furthermore, the claimed subject matter is not limited to implementations that address any disadvantages noted above or in any part of the present disclosure. Short description of the drawings Fig. 1 shows a schematic diagram of an exemplary engine system. Fig. Figure 2 shows a map showing the frequency content of a cylinder at idle under different degrees of imbalance in a prime mover, such as the prime mover system of Fig. 1, shows. Fig. 3 illustrates an exemplary flowchart illustrating a routine for determining a cylinder with an air / fuel imbalance in an engine, such as the engine system of Fig. 1, illustrated. Fig. Figure 4 shows an example torque model. Fig. 5 shows a schematic diagram of an example of engine speed signal processing using a notch filter. Fig. Figure 6 shows an example map of the frequency response of the notch filter. Fig. Figure 7 shows an example comparison map of the filtered engine speed signal at different degrees of lean imbalance. Fig. Figure 8 shows an example of the normalization process. Fig. Figure 9 shows an example of two PIP samples per ignition cycle. Fig. Figure 10 shows an example flowchart illustrating the signal normalization process. Detailed description

[0011] The following description relates to systems and methods for detecting a potential air / fuel imbalance in an engine, such as the engine of Fig. 1. An example map showing a comparison of the frequency domain characterization using a discrete Fourier transform of engine speed under engine idle conditions at stoichiometry and different degrees of lean imbalance is shown in Fig. 2. An engine control unit may be configured to implement a control routine for monitoring air / fuel imbalance, such as Fig. 3 and Fig. 5, to perform a frequency-domain characterization of the engine speed with respect to changes in crankshaft accelerations via a discrete notch filter in which only a narrow frequency band of adjacent frequencies or a single frequency is passed through the filter output. The output obtained using the notch filter can respond to changes at a selected multiple of the firing frequency, e.g., once per engine cycle (two engine revolutions for a 4-stroke engine) ( Fig. 6). Changes in the level of the filter output relative to that of a reference signal can be used to detect the presence and extent of cylinder imbalance ( Fig. 7). An example of signal processing of sampled data is shown in Fig. 8-9. Overall, the cylinder air / fuel imbalance can be accurately determined using a less time- and computationally intensive approach.

[0012] Fig. 1 shows a schematic diagram of an exemplary cylinder 30 in an internal combustion engine 10. The cylinder 30 may also be referred to herein as a combustion chamber 30. The engine 10 may be controlled at least in part by a control system including a controller 12 and by input from a vehicle operator 132 via an input device 130. In this example, the input device 130 includes an accelerator pedal and a pedal position sensor 134 for generating a proportional pedal position sensor signal PP.

[0013] The combustion chamber 30 of the engine 10 may have combustion chamber walls 33 with a piston 36 positioned therein. The piston 36 may be coupled to the crankshaft 40 such that reciprocating motion of the piston is translated into rotational motion of the crankshaft. The crankshaft 40 may be coupled to at least one drive wheel of a vehicle via an intermediate gear system (not shown). Furthermore, a starter motor may be coupled to the crankshaft 40 via a flywheel (not shown) to enable starting operation of the engine 10.

[0014] The combustion chamber 30 may receive intake air from the intake manifold 44 via the intake passage 42 and may exhaust combustion gases via the exhaust pipe 48 and the exhaust passage 58. The intake manifold 44 and the exhaust pipe 48 may be selectively connected to the combustion chamber 30 via an intake valve 52 and an exhaust valve 54, respectively. In some embodiments, the combustion chamber 30 may have two or more intake valves and / or two or more exhaust valves. In some examples, one or more of the intake passages may include a boosting device, such as a turbocharger or a supercharger. For example, Fig. 1 illustrates engine 10 configured with a turbocharger having a compressor 174 disposed between intake passages 42 and 44, and an exhaust turbine 176 disposed along exhaust passage 48. Compressor 174 may be at least partially powered by exhaust turbine 176 via a shaft 180, with the boosting device configured as a turbocharger. However, in other examples, such as when engine 10 is provided with a supercharger, exhaust turbine 176 may optionally be omitted, and compressor 174 may be driven by a mechanical input from an engine or the engine.

[0015] In the Fig. 1, the intake valve 52 and the exhaust valve 54 may be controlled by cam actuation via respective cam actuation systems 51 and 53. The cam actuation systems 51 and 53 may each control one or more cams on one or more camshafts (not shown in Fig. 1) and may utilize one or more of cam profile switching (CPS), variable cam timing (VCT), variable valve timing (VVS), and / or variable valve lift (VVL) systems that may be actuated by controller 12 to vary valve operation. The angular position of the intake and exhaust camshafts may be determined by position sensors 55 and 57, respectively. In alternative embodiments, intake valve 52 and / or exhaust valve 54 may be controlled by electric valve actuation. For example, cylinder 30 may alternatively have an intake valve controlled via electric valve actuation and an exhaust valve controlled via cam actuation, including CPS and / or VCT systems.

[0016] A fuel injector 66 is shown coupled directly to the combustion chamber 30 for injecting fuel directly therein in proportion to the pulse width of the FPW signal received from the controller 12 via an electronic driver 99. In this manner, the fuel injector 66 provides what is known as direct fuel injection into the combustion chamber 30. The fuel injector may be mounted, for example, in the side of the combustion chamber or in the top of the combustion chamber. Fuel may be delivered to the fuel injector 66 by a fuel system (not shown) including a fuel tank, a fuel pump, and a fuel rail.In some embodiments, the combustion chamber 30 may alternatively or additionally include a fuel injector disposed in the intake manifold 44 in a configuration that provides so-called intake port injection of fuel into the intake port upstream of the combustion chamber 30.

[0017] Ignition system 88 may provide an ignition spark to combustion chamber 30 via spark plug 91 in response to the spark advance signal SA from controller 12 during selected operating modes. Although spark ignition components are shown, in some embodiments, combustion chamber 30 or one or more other combustion chambers of engine 10 may be operated in a compression ignition mode with or without an ignition spark.

[0018] The intake manifold 44 is shown connected to a throttle 62 having a throttle plate 64. In this particular example, the position of the throttle plate 64 may be varied by the controller 12 via a signal provided to an electric motor or actuator (not shown). Fig. 1) present in the throttle 62, a configuration commonly called electronic throttle control (ETC). The throttle position may be varied by the electric motor via a shaft. The throttle 62 may control airflow from the intake passage 42 to the intake manifold 44 and the combustion chamber 30 (and other engine cylinders). The position of the throttle plate 64 may be provided to the controller 12 by the throttle position signal TP from the throttle position sensor 158.

[0019] The exhaust gas sensor 126 is coupled to the exhaust manifold 48 upstream of an emission control device 70. The sensor 126 may be any suitable sensor for providing an indication of an exhaust air / fuel ratio, such as a linear oxygen sensor or UEGO (Universal or Wide-Range Exhaust Gas Oxygen) sensor, a dual-state oxygen sensor or EGO, a HEGO (Heated EGO), a NOx, HC, or CO sensor. The emissions control device 70 is shown disposed along the exhaust passage 58 downstream of the exhaust gas sensor 126. The device 70 may be a three-way catalyst (TWC), a NOx trap, various other emissions control devices, or combinations thereof.

[0020] An exhaust gas recirculation (EGR) system (not shown) may be used to direct a desired portion of exhaust gas from the exhaust passage 58 to the intake manifold 44. Alternatively, a portion of combustion gases may be retained in the combustion chambers as internal EGR by controlling the timing of the exhaust and intake valves.

[0021] The control 12 is in Fig. 1 as a conventional microcomputer including a microprocessor unit 102, input / output ports 104, read-only memory 106, random access memory 108, maintain memory 110, and a conventional data bus. The controller 12 controls various actuators, such as a canister purge valve (not shown), the throttle plate 64, the fuel injector 66, and the like.Controller 12 is shown receiving, in addition to the previously discussed signals, various signals from sensors coupled to engine 10, including engine coolant temperature (ECT) from temperature sensor 112 coupled to cooling sleeve 114; a position sensor 134 coupled to an accelerator pedal 130 for sensing the accelerator pedal position set by the vehicle operator 132; an intake manifold pressure (MAP) measurement from pressure sensor 121 coupled to intake manifold 44; a profile ignition pickup (PIP) signal from a Hall effect sensor 118 (or other type of sensor) coupled to crankshaft 40; a measurement of air mass entering the engine from airflow sensor 120; a measurement of throttle position from sensor 158; and an air / fuel ratio (AFR) from EGO sensor 126.In a preferred aspect of the present description, crankshaft sensor 118, which may be used as an engine speed sensor, may produce a predetermined number of evenly spaced pulses during each revolution of the crankshaft, from which engine speed (RPM) may be determined. Such pulses may be transmitted to controller 12 as a profile ignition pickup (PIP) signal, as previously mentioned. Crankshaft sensor 118 may also be used to measure crankshaft accelerations (also called crank accelerations).

[0022] The read-only memory storage medium 106 may be programmed with computer-readable data representing instructions executable by the processor 106 to perform various routines not specifically listed herein. The controller 12 thus receives signals from the various sensors of Fig. 1 and uses the various actuators of Fig. 1 for adjusting the engine operation based on the received signals and instructions stored in a memory of the controller.

[0023] As previously described, Fig. 1 depicts only one cylinder of a multi-cylinder engine, and that each cylinder has its own group of intake / exhaust valves, fuel injectors, spark plugs, etc. In one example, engine 10 may include four cylinders configured in a linear manner. In another example, engine 10 may include six cylinders configured in a V-configuration. In yet another example, engine 10 may include eight cylinders configured in a V-configuration. Alternatively, engine 10 may include additional or fewer cylinders without departing from the scope of this disclosure.

[0024] A controller, such as controller 12, of the engine may monitor the cylinder air / fuel ratio imbalance at regular intervals to detect an air / fuel ratio change between cylinders. As such, air / fuel ratio imbalances within cylinders may negatively impact engine performance and engine emissions. As discussed herein with respect to Fig. 3-7, air / fuel imbalances can be detected by analyzing frequency harmonics of the engine speed using a signal processing notch filter.

[0025] Therefore, the presence of an air / fuel ratio imbalance in a cylinder can alter the frequency spectrum of the instantaneous engine speed. The average engine speed can remain close to the reference speed for both balanced and unbalanced cylinders due to the fueling control system. In the case of a healthy engine (where all cylinders are balanced), there is smooth torque production with a very small amplitude of low-frequency harmonics. However, in the case of an unhealthy engine (where one or more cylinders may be unbalanced), uneven torque production causes jumps in the low-frequency harmonic spectrum.The controller may determine a cylinder with an air / fuel imbalance in the engine by applying the notch filter to analyze the frequency spectrum produced by each cylinder. Specifically, the notch filter may be applied at the sampling frequency. Based on the notch-filtered frequency spectrum, which differs in magnitude from a notch-filtered reference spectrum of a balanced cylinder, the controller may determine that an imbalance condition exists and take appropriate mitigating measures.

[0026] Now on Fig. 2, this shows changes in the frequency spectrum of the instantaneous engine speed for an engine cylinder. In particular, Fig. Two maps 200-206, depicting the change in the frequency content of an engine cylinder at idle under varying degrees of imbalance. The idle engine speed content of the engine cylinder is analyzed using the Discrete Fourier Transform (hereafter referred to as DFT). The DFT converts a sampled function from its original domain (usually time or position along a line) to a frequency domain. Fig. Figure 2 shows the coefficients of Fourier series for four different scenarios.

[0027] The first map 200 shows the frequency content of a cylinder at 0% imbalance. In other words, map 200 represents a healthy cylinder. Therefore, in a healthy engine, where all cylinders have 0% imbalance, there is uniform torque production, which translates into a very small amplitude in low-frequency harmonics. However, when an engine is unbalanced, there is uneven torque production, resulting in low-frequency harmonics that jump across the frequency spectrum, with the strength of the low-frequency harmonics increasing with increasing cylinder imbalance.This is shown by the circled area in maps 200-206, where map 202 represents low frequency harmonics in a cylinder with a 10% lean imbalance, map 204 represents low frequency harmonics in a cylinder with a 20% lean imbalance, and map 206 represents low frequency harmonics in a cylinder with a 30% lean imbalance.

[0028] While DFT-based frequency analysis can be used to detect cylinder imbalances, relying on DFT processing makes the analysis time-consuming, expensive, and complex. As elaborated herein, by using discrete notch filtering of the engine speed content at a frequency corresponding to the sampling frequency, low-frequency harmonics can be detected in a more cost-, time-, and computationally efficient manner.

[0029] Fig. 3 shows an exemplary method 300 for detecting cylinder air / fuel imbalances in an engine using a discrete notch filter. The method 300 is used in conjunction with the Fig. 1, however, it should be understood that similar routines may be used with other systems without departing from the scope of this disclosure. Instructions for carrying out method 300, as well as other methods included herein, may be executed by a controller, such as the controller 12 of the Fig. 1, based on instructions stored in a memory of the controller and in conjunction with signals received from sensors of the engine system, such as the sensors described above with reference to Fig. 1. The controller can control engine actuators of the engine system, such as the actuators of the Fig. 1, to adjust engine operation and vehicle operation according to the routines described below.

[0030] At 302, method 300 includes estimating and / or measuring engine operating conditions. Example engine operating conditions evaluated include engine speed (Ne), engine load, torque request, MAP, ambient conditions (temperature, pressure, humidity, etc.), spark timing, boost pressure, etc. For example, the controller may receive signals from a MAP sensor, such as MAP sensor 121. Fig. 1, to determine an existing intake manifold pressure.

[0031] Next, at 304, the method includes determining whether input conditions for enabling air / fuel ratio imbalance monitoring (also referred to herein as air / fuel monitoring) are met. Various input conditions for starting air / fuel monitoring may be checked at 304. In one example, the input conditions may be confirmed when a background sampling rate (e.g., time-based sampling) and / or a foreground sampling rate (e.g., crank angle range-based sampling) is higher than a threshold rate. In another example, the input conditions may depend on engine operating conditions. For example, the input conditions may be confirmed in response to one or more engine temperatures being warmer than a threshold temperature, engine load being less than a threshold load, engine speed being less than a threshold speed, etc.In yet another example, air / fuel imbalance monitoring may be enabled only in response to steady-state engine conditions, and monitoring may be disabled in response to transient engine conditions, such as tip-ins and tip-outs. In still further examples, monitoring may be performed according to a predefined schedule, such as once per threshold distance of vehicle travel, once per threshold duration of vehicle travel, once per drive cycle, etc.

[0032] If the input conditions are not met, the method includes delaying activation of the monitor at 306. That is, the controller may wait to activate the air / fuel imbalance monitor while nominal engine operation continues. The AFR imbalance routine then ends. The routine may be deactivated and rescheduled for a later time, e.g., after a certain number of miles have been driven, after a certain period of time has elapsed, steady-state engine conditions, etc.

[0033] However, if the input conditions are met, method 300 proceeds to enable air / fuel ratio imbalance monitoring to estimate AFR in each cylinder of the engine. Thus, if the input conditions are met, method 300 includes, at 308, sampling engine speed values ​​multiple times within an engine firing event. In particular, the engine speed values ​​may be sampled concurrently with engine firing events. For example, engine speed may be sampled once per firing event. Alternatively, engine speed may be sampled at an integer multiple of the firing frequency. In one example, a crankshaft sensor that monitors the rotational speed of the crankshaft may send PIP signals to the controller.Based on the PIP signals received from the crankshaft sensor, the controller may generate engine speed signals that may be sampled during an engine firing event. The sampling frequency may be based on the engine configuration and firing frequency. For example, in an inline 4-cylinder engine, the engine speed may be sampled four times per engine cycle.

[0034] The sampling frequency can be the frequency or timing allowed by specific hardware. For example, it can be once per ignition cycle sampling or twice per ignition cycle sampling (respectively called pip or half-pip sampling). Sampling can be more frequent if the hardware has a specific tooth-synchronized signal. The notch filter output is determined for nominal hardware (limit part specification, for example, 7% lean or rich) and then determined at the emissions error hardware (near 20% lean or rich), the latter of which defines the threshold for the notch filter.

[0035] At 310, prior to further processing the engine speed values, the method includes normalizing the engine speed value of each cylinder with respect to changes in crankshaft angle. As with respect to Fig. 8, discrete crankshaft acceleration (or torque) calculations are derived from a half-pip timer. The controller uses logic to first calculate the crankshaft speed and then the crankshaft acceleration. Therefore, three time-stamped events are necessary for a final crankshaft acceleration / torque calculation in a 4-cylinder engine. A normalization process then follows, adjusting a known variation to equalize the magnitude of the first harmonic with known injectors and a predetermined imbalance level. As with respect to Fig. 10, during the normalization process, engine torque changes are first calculated using engine speed values ​​derived from crankshaft sensor gear timing snapshots. Then, the values ​​are normalized based on spark advance and torque demand at the given engine speed load.

[0036] At 312, method 300 includes applying a notch filter to the sampled data at a specified frequency. The specified frequency may correspond to the sampling frequency. In one example, filtering includes setting the notch filter to a frequency of once per engine cycle. Thus, a discrete notch filter is applied at a frequency to extract various components of the engine speed frequency content. The output is then referred to as the notch filter output. A transfer function of the filter in the crank angle domain may be given as follows: G(zθ)=Kf1−2. cos ωn.zθ−1+zθ−21−2.r. cos ωn.zθ−1+r2.zθ−2 where G is the discrete time domain transfer function of the notch filter, K f the reinforcement is, e.g. θ is the delay operator and ω is the frequency of interest. ωn=2πfsfn Kf={1if fn=01−2.r.cos(ωn)+r22−2.cos(ωn)if fn≠0 f s =12 f n =0,1,..6 r=0.99

[0037] For a 6-cylinder engine, fs (sampling frequency) is 12 (considering 12 samples per engine cycle). fn is the frequency of interest.

[0038] At 314, the method includes subtracting the notch filter output from the originally sampled signal to obtain a frequency content of the cylinder, specifically a frequency of a wave of interest. The resulting signal is the first sine term in the Fourier series of the engine speed signal and is defined as follows: M1(θ) = |M1| sin(θ+φ)

[0039] At 316, the frequency content of the cylinder is compared to a threshold, where the threshold is based on the frequency content of a cylinder when no imbalance is present. For example, a magnitude of the frequency content may be compared to a threshold. Based on the comparison, it is determined whether the frequency content indicates that a cylinder imbalance is present. In particular, the cylinder imbalance results in the first component of the frequency content being non-zero (compared to that of a healthy engine), and the wave analysis corresponding to this frequency is used for fault detection (as with respect to Fig. 7). Furthermore, an orientation of the imbalance can be determined based on the magnitude and phase of the frequency content. In one example, if the frequency content is greater than the threshold, it can indicate that an imbalance condition exists.

[0040] If an imbalance is not determined, the routine proceeds to 318, where no cylinder imbalance is indicated, such as by maintaining the clearing of an imbalance flag. Additionally, at 320, in response to an indication of no cylinder imbalance, the engine air and fuel parameters are maintained.

[0041] In comparison, at 322, a cylinder imbalance is indicated, such as by setting a diagnostic code or flag when a cylinder imbalance is determined. Additionally, at 322, engine air / fuel parameters may be adjusted based on the cylinder imbalance indication to correct the imbalance.

[0042] Now referring to Fig. 10, an exemplary method 1000 for normalizing sampled signals is described. In one example, the method 1000 may be implemented as part of the routine of Fig. 3, such as step 310, such as once per PIP.

[0043] At 1002, the method includes enabling the air-fuel monitor. From there, the method proceeds to normalize the sampled signals in a first approach based on the spark deviation, as detailed at 1004-1010, and normalize the sampled signals in a second approach based on the gear profile, as detailed at 1014-1026. Therefore, both normalization approaches are performed concurrently.

[0044] At 1004, the method includes estimating the acceleration during the power stroke per firing cylinder. In particular, the estimated acceleration is based on a position of a timestamp corresponding to the sampled signal within the power stroke (e.g., based on whether the signal was sampled at the beginning, middle, or end of the power stroke). For example, torque is determined as a product of a scalar and the estimated acceleration. The final acceleration value is calculated per half-pip, and then one sample later (another half-pip later), the required synchronization, if it lies in the appropriate cell, is linked to the estimated cylinder acceleration.

[0045] At 1006, the method includes normalizing the estimated acceleration by the value of the specified torque less any auxiliary loads. At 1008, the estimated scaled acceleration is normalized by the offset between the spark at MBT and the spark advance employed for the cylinder (i.e., based on the cylinder's spark ratio). This includes applying a numerator at the normalized torque close to 1 under ideal conditions (operating at stoichiometry and with the spark at MBT). For example, control may calculate the delta between the MBT and the actual spark advance for a given cylinder. For non-split injection of fuel into the cylinder, control determines the correction coefficient using a predetermined function lookup table. Control then calculates the normalized final torque estimate. The final scalar normalizer is pre-calibrated.The value of the scalar normalizer is determined under ideal MBT conditions such that the ratio of the estimated acceleration and tq_ind_led (specified torque based on the air charge at optimal spark and with lambda = 1) is close to 1. At 1010, the attributes of the completed patterns are stored. In particular, the determined values ​​for the normalized torque are stored for each cylinder / pattern. Therefore, it is also known that the two primary factors that can affect the timing input to the crankshaft acceleration calculation are: (i) cylinder position relative to the crankshaft and crankshaft position sensor (CKP) position, also known as torque variations, and (ii) gear "profile," such as geometry variations and the like.For these two factors, the experimental data is sampled with the actual hardware to obtain the deviation of the frequency signal from the ideal harmonics at the firing frequency. The final values ​​are corrected using an experimentally determined correction factor. These parameters can influence the magnitude of the [M] component of the first harmonic, which is defined by the notch filter as M1(θ) = [M1]Sin(θ+φ).

[0046] Thus, at 1014, the method includes sampling the crankshaft speed input. At 1016, the method includes caching the input. The caching may include the number of samples necessary to calculate the frequency components based on the equation previously analyzed on pages 13-14. In one example, three samples are needed: a current sample, a pre-sample, and a pre-pre-sample. At 1024, the method proceeds to extract the frequency components based on the cached data.

[0047] From each of steps 1024 and 1010, the method proceeds to 1026, where the compensated (normalized) output is further processed (and the routine returns to Fig. 3, such as step 312).

[0048] With reference now to Fig. 4, map 400 illustrates an exemplary torque model overview. Torque modeling includes the determination of an indicated torque (tq_ind_led) by a controller K. The indicated torque is derived from an air charge estimate assuming optimal spark timing (at MBT) and optimal lambda (of 1 or stoichiometry), with all cylinders of the engine active and fueled.

[0049] With reference now to Fig. 5 shows a block diagram 500 illustrating the signal processing of an engine speed signal by a notch filter. The signal processing enables evaluation of the low-frequency harmonics of the signal. A jump in the first component of the frequency spectrum (i.e., the frequency one period per engine cycle) is used to infer cylinder torque imbalance. The engine speed signal 502 is passed through a discrete notch filter 504 to obtain a processed signal or notch filter output 506. By subtracting the processed signal 506 from the engine speed signal 502, which is the original signal, a filtered component is obtained that may represent the first frequency content of the engine speed. By comparing this first frequency content to a reference or threshold value, the cylinder imbalance is determined.

[0050] With reference now to Fig. In Figure 6, map 600 illustrates an example variation in the frequency response of a notch filter. The upper graph (plot 602) represents the magnitude of the frequency response, while the lower graph (plot 604) represents the phase. The x-axis shows the normalized frequency with the response to the firing frequency. The filter sweeps through all frequencies unchanged except for the frequency of interest (the first frequency).

[0051] With reference now to Fig. 7, map 700 illustrates an exemplary change in the magnitude of the frequency spectrum of an engine speed signal with a change in the degree of imbalance. As can be seen by comparing the graph for a healthy engine with no imbalance (solid line) with those of engines with increasing degrees of imbalance (see dashed or dotted line graphs), the first frequency content of the wave jumps with increasing imbalance. Specifically, the balanced engine exhibits a first zero frequency content, while the unbalanced engines exhibit a first non-zero frequency content, with the non-zero content increasing with increasing engine imbalance (lean or rich; lean is depicted in the given example).

[0052] With reference now to Fig. 8, map 800 illustrates the process of normalizing or adjusting the change in the engine speed signal to equalize the magnitude of the first harmonic with known injectors and a predetermined imbalance level. Cylinder firing is denoted by Firing_inj_1 and 2. Timestamps are represented as TS_1, TS_2, TS_3. The controller may determine the engine speed for each firing event as follows: Speed_1=(TS_2−TS_1) / ΔTime; Speed_2=(TS_3−TS_2) / ΔTime; Acceleration=(speed_1−speed_2) / Δtime; TQ~Acceleration;

[0053] The acceleration occurs 3 samples later. Sync_ctr_0 is used to synchronize the acceleration with the firing.

[0054] Fig. Figure 9 shows the so-called half-PIP sampling on the 900 card. In this case, the signal is sampled at the up / down interrupt. Fig. Figure 9 also illustrates the limitation of sampling in response to synchronization with MBT. A small phase shift in synchronization results in a large difference in acceleration, which can impact the final torque estimate. This highlights the need for faster sampling than half-PIP sampling. In particular, the first half of the curve (ascending portion) shows a pip-up transition, while the second half of the curve (descending portion) shows a pip-down transition. The torque acceleration snapshot for both transitions illustrates the problems associated with two PIP samples for one ignition cycle.

[0055] As a result, the occurrence of low-frequency harmonics in a notch-filtered engine speed signal that jump in the first component of the frequency spectrum can be advantageously used to detect torque imbalance between cylinders. By then using the attributes of the low-frequency harmonics (including their strength, amplitude, etc.) to determine the magnitude and nature of the imbalance, appropriate mitigation measures can be implemented. By relying on a notch-filtered output, the time and cost associated with detecting torque imbalances between cylinders are reduced. Therefore, the use of the notch filter allows for simplified normalization, reducing computation time and cost.Additionally, the air / fuel ratio imbalance can be reliably determined without requiring computational complexity. Overall, the technical effect of determining cylinder AFR with a higher degree of accuracy enables adjustments to engine operation based on the detected AFR imbalance. Accordingly, engine performance can be improved and emissions reduced.

[0056] In one example, a method includes sampling an engine speed signal coincident with engine firing; filtering the sampled engine speed signal via a notch filter; and indicating a cylinder air / fuel imbalance based on an output of the notch filter relative to a threshold. In the previous example, filtering additionally or optionally includes setting the notch filter to a frequency of once per engine cycle. In any of the previous examples, indicating additionally or optionally includes indicating a magnitude of the imbalance based on a magnitude of a deviation of the notch filter output relative to the threshold.In any or all of the preceding examples, indicating the cylinder air / fuel imbalance additionally or optionally comprises indicating a deviation of an air / fuel ratio of the cylinder from a predetermined air / fuel ratio, wherein the predetermined air / fuel ratio comprises stoichiometry. In any or all of the preceding examples, the method additionally or optionally further comprises, prior to filtering, excluding engine speed signal variations of the crankshaft, wherein the excluded engine speed signals are not used to indicate the cylinder air / fuel imbalance. In any or all of the preceding examples, the method additionally or optionally further comprises adjusting one or more engine operating parameters in response to indicating a cylinder air / fuel imbalance.In any or all of the preceding examples, sampling the engine speed signal additionally or optionally comprises sampling an output of a crankshaft sensor coupled to a crankshaft, the method further comprising correcting the sampled engine speed signal based on a learned gear profile of the crankshaft. In any or all of the preceding examples, additionally or optionally, the gear profile of the crankshaft is learned once during initial engine operation and then maintained during subsequent engine operation. In any or all of the preceding examples, the notch filter additionally or optionally has an infinite impulse response, and wherein a frequency of the notch filter is based on a frequency of the sampling.In any or all of the preceding examples, sampling engine speed signals for a given engine cylinder additionally or optionally comprises sampling the engine speed signal during a power stroke of the cylinder, wherein the method, prior to filtering, further comprises normalizing the sampled engine speed signal for the given engine cylinder based on a sampling time within the power stroke. In any or all of the preceding examples, the normalizing additionally or optionally comprises reducing the torque of the given engine cylinder based on a deviation of the ignition timing from the MBT spark at the sampling time within the power stroke.

[0057] Another example method for an engine includes: sampling crankshaft acceleration during a power stroke of a firing cylinder; normalizing the sampled acceleration based on a time of sampling within the power stroke; filtering the normalized sampled acceleration through a band-stop filter; and, if noise at the filter is lower, indicating a cylinder air / fuel imbalance based on the output of the band-stop filter relative to a threshold. In the previous example, if noise at the filter is higher, the method additionally or optionally further includes indicating a cylinder air / fuel imbalance based on an exhaust air-fuel ratio sensor while disregarding the output of the band-stop filter.In any or all of the preceding examples, the method additionally or optionally further comprises adjusting an engine operating parameter based on the cylinder air / fuel imbalance. In any or all of the preceding examples, the method additionally or optionally further comprises adjusting a parameter of the notch filter based on the normalization, wherein the notch filter comprises a notch filter, the parameter comprising a bandwidth and a band frequency of the notch filter. In any or all of the preceding examples, the adjusting additionally or optionally comprises broadening the band of the notch filter as a degree of normalization increases.In any or all of the previous examples, normalizing additionally or optionally includes correcting the sampled acceleration based on geometry changes corresponding to a learned crankshaft gear profile; and reducing the torque of the given engine cylinder based on a deviation of the ignition timing from the MBT spark at the time of sampling.

[0058] Another example engine system includes: an engine having a plurality of cylinders; a fuel injector for supplying fuel to at least one of the plurality of cylinders; an intake throttle coupled to an intake passage of the engine; a crankshaft sensor connected to a crankshaft gear; and a controller.The controller may be configured with computer-readable instructions stored in non-transitory memory to: sample an output of the crankshaft sensor concurrent with a cylinder firing event; correct the sampled output based on a learned gear profile; normalize the corrected sampled output based on a timing of the sampling relative to a stroke of the cylinder firing event; filter the corrected output via a notch filter, wherein a parameter of the notch filter is based on a frequency of the sampling; estimate a cylinder air / fuel imbalance based on the filtering; and adjust one or more of the intake throttle and the fuel injector based on the estimated cylinder air / fuel imbalance.In the previous example, normalizing additionally or optionally includes estimating engine torque based on engine acceleration corresponding to the corrected sampled output; estimating spark timing at the time of sampling; and decreasing the estimated engine torque when the spark timing at the time of sampling deviates from the MBT spark, wherein the deviation is based on the timing of sampling relative to the power stroke, wherein the deviation is larger when the timing of sampling is at a beginning or end of the power stroke, wherein the deviation is smaller when the timing of sampling is at a middle of the power stroke.In any or all of the preceding examples, the controller additionally or optionally comprises further instructions for adaptively setting a bandwidth of the notch filter based on the normalization, wherein the bandwidth is widened when the deviation is larger.

[0059] It should be appreciated that the exemplary control and estimation routines included herein may be used with various engine and / or vehicle system configurations. The control methods and routines disclosed herein may be stored as executable instructions in non-transitory memory and may be executed by the control system comprising the controller combined with the various sensors, actuators, and other engine hardware. The specific routines described herein may represent one or more of any number of processing strategies, such as event-driven, interrupt-driven, multitasking, multithreading, and the like. Therefore, various actions, operations, and / or functions may be performed in the illustrated order, in parallel, or in some cases, omitted.Likewise, the order of processing is not necessarily required to achieve the features and advantages of the embodiments described herein, but is provided for ease of illustration and description. One or more of the illustrated actions, operations, and / or functions may be performed repeatedly depending on the particular strategy employed. Furthermore, the described actions, operations, and / or functions may graphically represent code to be programmed into non-transitory memory of the computer-readable storage medium in the engine control system, wherein the described actions are performed by executing the instructions in a system including the various engine hardware components in combination with the electronic controller.

[0060] It is understood that the configurations and routines disclosed herein are exemplary in nature, and that these specific embodiments are not to be considered limiting, as numerous variations are possible. For example, the above technology may be applied to V-6, I-4, I-6, V-12, horizontally opposed four, and other engine types. The subject matter of the present disclosure includes all novel and non-obvious combinations and sub-combinations of the various systems and configurations and other features, functions, and / or characteristics disclosed herein.

[0061] The following claims particularly point out certain combinations and subcombinations that are considered novel and non-obvious. These claims may refer to "a" element or "a first" element, or the equivalent thereof. Such claims should be understood to encompass the inclusion of one or more such elements, neither requiring nor excluding two or more such elements. Other combinations and subcombinations of the disclosed features, functions, elements, and / or properties may be claimed by amending the present claims or by presenting new claims in this or a related application. Such claims, whether broader, narrower, the same, or different in scope from the original claims, are also considered to be included within the subject matter of the present disclosure.

Claims

[1] Procedure comprising: Sampling an engine speed signal concurrent with engine firing; Filtering the sampled engine speed signal via a notch filter; and Indicating a cylinder air / fuel imbalance based on an output of the notch filter with respect to a threshold value, further comprising adjusting one or more engine operating parameters in response to the indication of the cylinder air / fuel imbalance. [2] The method of claim 1, wherein filtering comprises setting the notch filter to a frequency of once per engine cycle. [3] The method of claim 1, wherein the indicating comprises indicating a magnitude of the imbalance based on a magnitude of a deviation of the notch filter output with respect to the threshold. [4] The method of claim 1, wherein indicating the cylinder air / fuel imbalance comprises indicating a deviation of an air / fuel ratio of the cylinder from a predetermined air / fuel ratio, wherein the predetermined air / fuel ratio comprises stoichiometry. [5] The method of claim 1, further comprising, prior to filtering, excluding crankshaft engine speed signal variations, wherein the excluded engine speed signals are not used to indicate the cylinder air / fuel imbalance. [6] The method of claim 1, wherein sampling the engine speed signal comprises sampling an output of a crankshaft sensor coupled to a crankshaft, the method further comprising correcting the sampled engine speed signal based on a learned gear profile of the crankshaft, the corrected sampled engine speed then being filtered via the notch filter. [7] The method of claim 6, wherein the crankshaft gear profile is learned once during initial engine operation and then maintained during subsequent engine operation. [8] The method of claim 1, wherein the notch filter has an infinite impulse response, and wherein a notch frequency of the notch filter is based on a frequency of the sampling. [9] The method of claim 1, wherein sampling engine speed signals for a given engine cylinder comprises sampling the engine speed signal during a power stroke of the cylinder, the method further comprising, prior to filtering, normalizing the sampled engine speed signal for the given engine cylinder based on a time of sampling within the power stroke, the normalized engine speed then being filtered via the notch filter. [10] The method of claim 9, wherein normalizing comprises reducing the torque of the given engine cylinder based on a deviation of the ignition timing from the MBT spark at the time of sampling within the power stroke. [11] A method for a prime mover comprising: Sampling the crankshaft acceleration during a power stroke of a firing cylinder; Normalizing the sampled acceleration based on a sampling time within the working cycle; Filtering the normalized sampled acceleration through a band-stop filter; and when the noise at the filter is lower; indicating the cylinder air / fuel imbalance based on the output of the band-stop filter with respect to a threshold value, further comprising adjusting an engine operating parameter based on the indicated cylinder air / fuel imbalance. [12] The method of claim 11, further comprising, when the noise at the filter is higher, indicating a cylinder air / fuel imbalance based on an exhaust air / fuel ratio sensor while disregarding the output of the band-stop filter. [13] The method of claim 11, further comprising adjusting a parameter of the band-stop filter based on the normalization, wherein the band-stop filter comprises a notch filter, the parameter comprising a bandwidth and a band frequency of the band-stop filter. [14] The method of claim 13, wherein adjusting comprises broadening the band of the band-stop filter as a degree of normalization increases. [15] The method of claim 11, wherein normalizing comprises: Correcting the sensed acceleration based on geometry changes according to a learned crankshaft gear profile; and Reducing the torque of the given engine cylinder based on the deviation of the ignition timing from the MBT spark at the time of sampling. [16] Power machine system comprising: an engine having multiple cylinders; a fuel injector for supplying fuel to at least one of the plurality of cylinders; an intake throttle coupled to an intake passage of the engine; a crankshaft sensor coupled to a crankshaft gear; and a controller with computer-readable instructions stored in non-volatile memory for the following: Sampling a crankshaft sensor output concurrent with a cylinder firing event; Correcting the sampled output based on a learned gear profile; Normalizing the corrected sampled output based on a sampling time relative to a power stroke of the cylinder firing event; filtering the corrected output through a notch filter, wherein a parameter of the notch filter is based on a sampling frequency; Estimating a cylinder air / fuel imbalance based on filtering; and Adjusting one or more of the intake throttle and fuel injector based on the estimated cylinder air / fuel imbalance. [17] The engine system of claim 16, wherein normalizing comprises: Estimating engine torque based on the Engine acceleration according to the corrected sampled output; Estimating an ignition timing at the time of sampling; and Decreasing the estimated engine torque when the ignition timing at the sampling time deviates from the MBT spark, the deviation being based on the sampling time relative to the power stroke, the deviation being larger when the sampling time is at a beginning or end of the power stroke, the deviation being smaller when the sampling time is at a middle of the power stroke. [18] The engine system of claim 17, wherein the controller comprises further instructions for adaptively adjusting a bandwidth of the notch filter based on the normalizing, wherein the bandwidth is widened when the deviation is larger.

Images